problem with microdem "set Map pixel size"

problem with microdem "set Map pixel size" - Printable Version

+- All-Aircraft-Simulations (https://allaircraftsimulations.com)
+-- Forum: IL2 MAP MODS Download & Discussion (https://allaircraftsimulations.com/forumdisplay.php?fid=269)
+--- Forum: OLD IL2 MAP MODS Discussion & WIP-MAP-MODS (https://allaircraftsimulations.com/forumdisplay.php?fid=307)
+--- Thread: problem with microdem "set Map pixel size" (/showthread.php?tid=55607)

problem with microdem "set Map pixel size" - lowfighter - 05.08.2008

I just had a go at microdem, and followed ClockWatcher's tutorial, however no mater how much I set the map pixel size when I save I get always the same map.
For example I set size to 200 and when saved I get a bitmap of 720 by 720
If I set size to 50 when saved I get again a bitmap of same size 720 by 720
So for some reason the set map pixel size is not working, any ideas?

- Monguse - 05.08.2008

Compare a known distance to pixels/meter on your map

I just exported a map using the 50 and 200 logic and I came out with a 1/8 scale map. SO I had to scale everything 800%

- lowfighter - 05.08.2008

Monguse thanks for the tip. I thought too about this way and measured using google earth etc. However if microdem is scaling to 1/8 a lot of elevation resolution is going in the proces, also when you scale back I think, so we're loosing height details in this way. So I just wonder what's up, you get the same scaling down with microdem as me, but reading the threads I see nobody mentioned this before?

Monguse Wrote:Compare a known distance to pixels/meter on your map

I just exported a map using the 50 and 200 logic and I came out with a 1/8 scale map. SO I had to scale everything 800%

- BBury - 05.08.2008

I had the same problem. No matter if I selected 50 or 200 they all came out the same size. 710px

What I discovered was that you can get a higher resolution (better quality) map by going into Microdem/Options and select the "map" tab.
You can reset the default resolution of the exported BMP's. They will still be too small for a 1:1 map and have to be resized but you will get a much finer map. I set mine to 4096x4096 and incxreased my memory to 96megs.

- lowfighter - 05.08.2008

Thanks!

- Skunkmeister - 05.08.2008

I found out by talking with Monguse yesterday that the DEM's for Europe/Med area have something funky going on with them.

I also discovered that it set everything in reverse proportion, in order to get 1pix=200m I had to set the pixel size to 0.005, which is 1/200.

- fly_zo - 05.08.2008

can this have something to do with SRTM data version ?.... i vaguely remember seeing some warnings about it when i searched for Cro SRTM files

Z

- BBury - 05.08.2008

I tired the .005 to get the 1:200 and it worked but the 1:50 still only gives the maximum set in microdem/options/maps default BMP size on the widest azimuth of the map.

- lowfighter - 06.08.2008

Hey thanks M8s for help! I rushed to check and setting the number smaller will get different output maps. However I found that a correct size was 0.0018 . I checked using google earth and the latitude and longitude "box" of the srtm data (just one srtm block). The vertical dimension measured on googleearth was 556 Km

The size of the image I got with 0.0018 scaling was 2764 by 2764 pixel. So the vertical dimension image was 2764*200=552800 or approx 553 Km

So the error was of about 3 Km in 550 Km...

EDIT:hm I don't quite get it, just realised on google that the srtm data is not a square so if I measured the horisontal dimension I got a much smaller distance (358 Km). However the microdem image is square so I don't undertsand, or maybe I understand that one has to correct the output image in one dimension (vertical or horisontal) too? Am I right? Then maybe it's now clear why I got the scale "right" with 0.0018 and you got it right with a different value?

- Redwulf__32 - 08.08.2008

lowfighter Wrote:Hey thanks M8s for help! I rushed to check and setting the number smaller will get different output maps. However I found that a correct size was 0.0018 . I checked using google earth and the latitude and longitude "box" of the srtm data (just one srtm block). The vertical dimension measured on googleearth was 556 Km

The size of the image I got with 0.0018 scaling was 2764 by 2764 pixel. So the vertical dimension image was 2764*200=552800 or approx 553 Km

So the error was of about 3 Km in 550 Km...

EDIT:hm I don't quite get it, just realised on google that the srtm data is not a square so if I measured the horisontal dimension I got a much smaller distance (358 Km). However the microdem image is square so I don't undertsand, or maybe I understand that one has to correct the output image in one dimension (vertical or horisontal) too? Am I right? Then maybe it's now clear why I got the scale "right" with 0.0018 and you got it right with a different value?

On the vertical correction issue.

I noticed the same thing while working my Denmark map, it came out seemingly "squashed". I figured that since the data most likely are based on views from GEO, a good approximate correction factor would be the mean latitude of the area in question. Tried correcting (or streching) my initial maps (_c and _h) for 56 degrees (stretch vertically be 1/cos(lat)). Voil

- lowfighter - 08.08.2008

Thanks Redwulf!

- Redwulf__32 - 09.08.2008

Thought you might like this......

Forging on with the 5th total rework of my Denmark map :roll: I decided to take a rather more firm approach to SRTM data and MicroDEM.

Found this site http://www.wasp.dk/Support/FAQ/Maps/SRTM2MAP.html
Followed the link to NASA's ftp server ftp://e0srp01u.ecs.nasa.gov/srtm/version2/

EDIT ----------------------
ftp link is the same as in ClockWatcher's tutorial, seems it just has been off-line and is now up again.
-----------------------------

As noted SRTM1 covers North America (Real Hi-res) while SRTM3 cover the rest of the globe (Hi-res)

Data is stored in zipped .hgt files and named using long/lat of lower left pixel.

.hgt files can be merged with MicroDEM and saved as a DEM file for further use.

The REALLY GOOD thing is ...... pixel size works as in ClockWatcher's tutorial and the map is projected correctly!

See scaled down (1:16) of the map_c, the detail is incredible in full scale!

[Image: Denmark_wip.jpg]

- lowfighter - 10.08.2008

Thanks again, trying to understand how it works too.
This data is sampled at different lat and long points, still don't understand if I have a data comprising say 5 degrees latitude and 5 degrees longitude, then this data is square as an image but if you translate the angles into distance it means that different pixels will have different distance equivalents. So I don't know yet how microdem handles that, also I don't know how il2 handle it's maps data (the world is round also in Olegworld hehe)

- BBury - 10.08.2008

I had to rescale my extracted map to fit the Olegworld dimensions. I scaled it using the vertical plane but it is only accurate in the north-south plane since that is what I used. East - west my map is off a good 10 percent. The only way around this would be to compress the east-west dimension as close as possible but I figure this would adversely effect the elevation profile for the "h" map. When I overlay satellite or other maps I have to stretch them out (east-west) to fit on the SRTM maps as well so those maps are also "flattened" to compensate for the earths curvature into 2D. This so the lat.-long. grids are not curved on paper.

I guess a map maker would have to decide which plane they want to size the map accurately depending on which way the most traveling will take place or the shape/size of the map.